Download View PDF - CiteSeerX

Journal of Personality and Social Psychology
1990, Vol. 58, No. 3, 397-408
Copyright 1990 by the American Psychological Association, Inc.
0022-3514/90/$00.75
Effects of Fortuitously Activated Constructs Versus Activated
Communication Goals on Person Impressions
Constantine Sedikides
University of Wisconsin--Madison
The relative impact of construct accessibility effects (i.e., the influence of accessible and applicable
constructs) and communication goal effects (i.e., the influence of constructs activated through communication goals) on both the interpretation and transmission of ambiguous target information and
Ss' private target impressions is examined. Whether communication goal effects can lead to recoding
of target information is also examined. Ss were primed with either positive or negative trait constructs, were presented with ambiguous target information, and were instructed to communicate
their description and impressions of the target to a recipient holding either positive, negative, or
neutral attitudes toward the target. Construct accessibility effects were present in the case of the
neutral recipient but were overridden by communication goal effects in the case of either the positive
or negative recipient. Furthermore, communication goals introduced at retrieval led to recoding of
target information.
with a stored construct. If the fit is good, the stimulus input is
interpreted and stored as an instance of the construct. However,
because social stimuli are often ambiguous and constructs lack
sharply defined boundaries, multiple constructs will frequently
be equally applicable for capturing the stimulus input. The
question then becomes which construct will be used for interpreting a particular behavior. This question has been addressed
by research on construct accessibility.
A typical construct accessibility experiment in the social psychological literature has two ostensibly unrelated parts. In the
first part of the experiment, subjects perform a priming task,
which is a manipulation intended to enhance the accessibility
of certain available constructs. For instance, subjects may reorganize jumbled sentences containing the critical trait constructs, or primes. (For examples of priming tasks, see Bargh
& Pietromonaco, 1982; Herr, 1986; Higgins, R_holes, & Jones,
1977; Martin, 1986; Smith & Branscombe, 1987; and Srull &
Wyer, 1979.) In the second part of the experiment, subjects process behavioral descriptions pertaining to a target person. Subjects' responses are then examined for evidence of categorization of target behaviors along the lines of accessible constructs.
The typical finding is that target behaviors are interpreted according to the constructs that are the most applicable to the
behaviors and the most accessible in memory at the time the
behaviors are encoded. In this article, this phenomenon is referred to as construct accessibility effects (after Smith & Branscombe, 1987, p. 162). The same phenomenon has also been
labeled contextual priming effects (e.g., Higgins & Stangor,
1988) and assimilation effects (e.g., Martin, 1986).
Two classes of models that have been developed to account
for construct accessibility effects are mechanistic and excitation
transmission models (see also Smith, 1984; Smith & Branscombe, 1987, 1988). Mechanistic models (Forbach, Stanners, &
Hochhaus, 1974; Wyer & Srull, 1981) use the metaphor of a
component part to represent constructs. Construct accessibility
effects are explained through the mechanistic movements of
component parts. Wyer and Srull's ( 198 l, 1986) bin model, for
Research on the psychological processes underlying the formation of person impressions has tacitly accepted the assumption that impression formation is governed by intraindividual
factors. The process of impression formation is thought to involve perceivers' attending to target behaviors and encoding the
behaviors in line with relevant accessible constructs, at least
when the behaviors are ambiguous. Thus, the impression of the
target is decisively shaped by the constructs that happen to be
accessible at encoding. (For alternative views, see Anderson,
1974; Brewer, 1988; and Jones & Davis, 1965.) However, the
impression formation process can also be influenced by interindividual factors. For instance, when impressions are formed
with a communicative intent, they may be molded by the demands of the communication process per se (e.g., characteristics of the recipient to whom target impressions are to be communicated), instead of or in addition to being shaped by relevant, accessible constructs. This article is concerned with the
joint influence ofintraindividual and interindividual factors on
the formation of person impressions.
Effects o f F o r t u i t o u s l y A c t i v a t e d C o n s t r u c t s
on Person I m p r e s s i o n s
Brunet (1957) maintained that impression formation involved the comparison of a stimulus input (e.g., a behavior)
This article is based on a doctoral dissertation completed at the Ohio
State University. I gratefully acknowledge the input of my dissertation
committee: Thomas Ostrom (committee chair), Neal Johnson, Clark
Leavitt, Richard Petty, and John Skowronski. I also acknowledge the
insightful comments Denis Hilton, Thomas Ostrom, Eliot Smith, Robert Wyer, and several anonymous reviewers provided on drafts of this
article, as well as the suggestions the Illinois Social Cognition Group
made at a presentation of parts of this work. James Be,ale, Jennifer Gerard, Alesia Holliday, Lee Metzler, Heide Newton, Mark Piechota, Jodi
White, and Roderick Young provided invaluable assistance with data
collection and coding.
Correspondence concerning this article should be addressed to Constantine Sedikides, Psychology Department, University of Wisconsin,
1202 West Johnson Street, Madison, Wisconsin 53706.
397
398
CONSTANTINE SED1KIDES
instance, postulates that constructs are stored in memory bins.
In any given bin, different constructs occupy different positions.
Copies of a construct may be made (through priming procedures) and then deposited in the bin on top of other constructs,
making the former more accessible and the latter less accessible.
The accessible construct will subsequently be used in the processing of relevant target information.
Excitation transmission models (Higgins & King, 1981; Reder, 1983) postulate that priming of a construct elevates its excitation level. When a construct is excited to a certain level, or
threshold, the construct is likely to be used in information processing. An accessible construct is a construct that has been excited to its threshold.
Effects o f Constructs Activated T h r o u g h
C o m m u n i c a t i o n Goals on Person Impressions
Construct Accessibility and Communication Goals
Communication, "the transmission of mental content" (Zajonc & Adelmann, 1987, p. 4), is an act of intense cognitive
restructuring. Zajonc (1960) showed that the mere anticipation
of communication causes changes in cognitive structures. Cognitive structures are reformatted to meet the demands of the
communicative situation. Such demands are for the production
of a grammatically and socially sound message and for the attainment of interpersonal goals.
Communication is the currency through which interpersonal
goals are achieved. According to McCann and Higgins (1988),
interpersonal goals include production goals (successfully managing a project), entertainment goals (sharing a leisure activity),
social reality goals (maintaining a common definition of social
reality), social relationship goals (reinforcing a rewarding social
relationship, avoiding conflict), and face goals (making a positive impression on others). For any of these goals to be reached,
communicators and recipients must coordinate the exchange of
their messages. Coordination can be considered the overarching
goal in interpersonal communication (Clark, 1985; Grice,
1975; Hilton & Slugoski, 1986).
Communicator-recipient coordination efforts have been
shown to play a crucial role in shaping the structure of the communicative message. For example, communicators tend to
modify their message as a function of whether the transmitted
information is new or old (Clark & Haviland, 1977; Higgins,
McCann, & Fondacaro, 1982; Strack, Martin, & Schwarz,
1988), and they tend to select either definite or indefinite pronouns to characterize the referent of the message, depending on
which class of pronouns is more helpful in recipient comprehension (Chafe, 1974). Communicators will change their position on an issue or their description of a target to be consistent
with their recipient's attitudes toward the target (Higgins &
Rholes, 1978; Manis, Corneli, & Moore, 1974; Newtson &
Czerlinsky, 1974). High authoritarians, as compared with low
authoritarians, will tailor their message to the attitudes of their
high-status recipient (Higgins & McCann, 1984). Finally, high
self-monitors, as contrasted with low self-monitors, will modify
their message to fit their recipients' attitudes (McCann & Hancock, 1983). Effects that result from coordination attempts between communicators and recipients are termed communication goal effects in this article. (Higgins & Stangor, 1988, labeled
such effects contextual adaptation effects.)
Recent work in social psychology has advocated a synergism
of social cognition and communication (Donohew, Sypber, &
Higgins, 1988; Higgins, 1981; Holtgraves, Srull, & Socall, 1989:
Kraut & Higgins, 1984; McCann & Higgins, 1988). This work
has pointed out the interpersonal nature of the impression formation process. Impressions are often formed in order to be
communicated. The formation, communication, and perseverance of impressions are bound to be influenced by communication goals.
In particular, communication goals may interfere with accessible constructs in the impression formation process. The interference can occur for a number of reasons. First, because of the
functional significance of communication goals in everyday life,
target information may tend to be encoded not in line with accessible and applicable constructs but, rather, in accord with
communication goals. Second, even if relevant target information tends to be encoded in concert with accessible constructs,
the encoding may not be final in anticipation of the upcoming
communication. As Zajonc (1960) demonstrated, cognitive
structures remain relatively open in anticipation of communication.
Theoretically, communication goals are also activated constructs. The potential interference of constructs activated
through communication goals with constructs activated
through fortuitous experiences can be described in the language
of the bin model as copies of communication-related constructs
moving above the fortuitously activated constructs in the relevant bin. In the language of excitation transmission models, the
excitation level of communication-related constructs may surpass the excitation level of fortuitously activated constructs.
The first objective of the present research was to examine
whether the effects of fortuitously activated constructs (construct accessibility effects) could be overridden by the effects of
constructs activated by means of communication goals (communication goal effects).
Can Communication Goals Lead to Recoding
of Target Information?
Construct accessibility effects are a function of encoding. In a
study by Srull and Wyer (1980), subjects completed the priming
task either before or after encoding target information. Construct accessibility effects were obtained only when the priming
task was completed before the encoding of target information.
In Srull and Wyer's (1980) words, "Once information has been
encoded in a particular way, it is typically not recoded in terms
of categories that are highly accessible at the time of judgment"
(p. 852).
However, Srull and Wyer (1980) qualified their remarks. They
suggested that caution should be exercised in concluding that
recoding will never occur. They stressed that "recoding is more
likely iftbe purposes for which the information is to be used are
introduced after the information is received and the original
encoding of the information is inadequate for these purposes"
(Srull & Wyer, 1980, p. 852).
A communicative situation satisfies the prescriptions for recoding set by Srull and Wyer (1980). If communication goals
are introduced after encoding of target information, initial encoding may prove to be inadequate for the fulfillment of communication goals (e.g., avoiding conflict with the recipient and
PERSON IMPRESSIONS
enabling the recipient to comprehend the message). Recoding
may be a reasonable alternative.
Thus, the second objective o f the present research was to find
out whether communication goal effects predominate over construct accessibility effects when c o m m u n i c a t i o n goals are introduced at the postencoding stage.
Communication Goals and Subjects'
Private Impressions
A substantial literature documents that subjects' previous
categorization o f target information, rather than the stored representation o f it, affects subsequent judgments (and memory)
o f the target (Carlston, 1980; Higgins & Lurie, 1983; Higgins &
Rholes, 1978). The third objective o f the present research was
to explore whether the impact of communication goals extends
to subjects' private social thinking.
Study 1
The first aim of Study 1 was to investigate whether construct
accessibility effects are moderated by communication goal
effects when the message recipient holds either favorable or unfavorable attitudes toward the referent of the message. Imagine
a situation in which a positive trait construct (e.g., frugal) has
become accessible in subjects' m e m o r i e s through a priming
technique. Subjects are next presented with ambiguous target
information (i.e., information that is equally likely to imply the
trait frugal or the trait stingy) and are asked to c o m m u n i c a t e
their impressions of the target to a recipient who dislikes the
target. Will subjects' descriptions and impressions be positively
toned (e.g.,frugal) or negatively toned (e.g., stingy)?
Consistent with established construct accessibility effects,
subjects' impressions should be affected by the applicable constructs that happened to be accessible at encoding. However,
consistent with established c o m m u n i c a t i o n goal effects, subjects' impressions should be affected by the evaluative direction
o f the recipient's attitudes.
The second aim of Study l was to test whether subjects' communicated impressions would in turn affect their own subsequent characterizations of the target.
399
in groups ranging in size from 5 to 8. Dividers on the tables in the experimental room prohibited subjects from seeing one another when seated.
Subjects were randomly assigned to conditions of a balanced factorial
design.
Experimental Design
Two target essays were constructed, one referring to Ralph and one
referring to Donald. Target of the essay constituted the first design variable. We included this replication variable to insure generality of the
results over targets.
Valence of primes constituted the second design variable. We included this variable to establish potential construct accessibility effects.
For half of the subjects, the primes were positive; for the other half of
the subjects, the primes were negative.
Applicability of primes was the third design variable. For half of the
subjects, the primes were applicable (i.e., subjects who were given the
Donald essay had also been given the Donald primes; subjects who were
given the Ralph essay had also been given the Ralph primes), whereas
for the remaining half of the subjects, the primes were inapplicable (i.e.,
subjects who were given the Donald essay had been given the Ralph
primes; subjects who were given the Ralph essay has been given the Donald primes). The applicability variable was included to show that exposure to positive or negative primes was not in itself sufficient to produce
construct accessibility effects. The primes had to be applicable to the
target information.
Order of the priming task and the communication instructions was
the fourth design variable. For half of the subjects, the priming task
came first, followed by communication instructions. For the remaining
half of the subjects, communication instructions preceded the priming
task. This variable was instituted to establish the generality of the potential moderation of communication goal effects on construct accessibility
effects. In other words, if communication goal effects moderate construct accessibility effects, they should do so regardless of which task is
temporally closest to the acquisition of target information or is more
accessible in subjects' memories.
Recipient of target message was the final design variable. A third of
the subjects were asked to communicate the essay to a recipient who
liked the target; another third were asked to communicate the essay to
a recipient who disliked the target; the rest of the subjects were asked to
communicate the essay to an unspecified recipient.
Thus, the design was a 2 (target: Donald, Ralph) × 2 (valence: positive
primes, negative primes) × 2 (applicability: applicable primes, inapplicable primes) × 2 (order of tasks: priming task first, communication
instructions first) × 3 (recipient: positive, negative, neutral) betweensubjects factorial.
Method
Overview
Subjects were primed with trait constructs that were either positive
or negative, and either applicable or inapplicable; were presented with
target information; and were asked to communicate in writing their descriptions and impressions of a target to a recipient who either liked,
disliked, or held neutral attitudes toward the target. The priming task
and the communication instructions were both given before encoding
of target information. However, for half of the subjects, the priming task
preceded communication instructions, whereas for the remaining subjects, communication instructions preceded the priming task. After
subjects communicated their impressions of the target, they listed their
own private impressions of the target.
Subjects
Subjects were 192 Ohio State University undergraduates participating for extra introductory psychology course credit. Subjects were tested
Construction of Target Essays
In constructing the target essays, I had to meet two objectives. First,
each paragraph of each essay had to be ambiguous, that is, equally likely
to be characterized by any of two (denotatively similar but evaluatively
different) traits. Second, the traits that were relevant to one essay had to
be irrelevant to the other essay.
The two essays had to describe the target in terms of behaviors that
exemplified connotatively similar but evaluatively different trait terms.
Bipolar adjective pairs having this characteristic were selected from two
sources. The first source was the Peabody (1967) monograph on the
descriptive and evaluative aspects of trait terms. The following five bipolar pairs of traits were selected (the numbers in parentheses correspond
to the evaluative rating given to the trait by Peabody's subjects on a scale
ranging from 3 to -3): moral (I.3) versus self-righteous (-1.6), thrifty
(0.9) versus stingy (-2.0), cultivated ( 1.6) versus artificial (-2.2), idealistic (1.5) versus unrealistic (-1.2), and witty (1.8) versus sarcastic
(-0.8). The second source of selection of bipolar adjective traits was
previous research on construct accessibility. The following three bipolar
400
CONSTANTINE SEDIKIDES
adjective pairs that had been pretested and used in past research were
adventurous versus reckless, independent versus aloof, and persistent
versus stubborn (Higgins & McCann, 1984; Higgins et al., 1977).
Next, eight paragraphs were constructed, corresponding to the eight
bipolar adjective pairs. Each paragraph exemplified both members of
the adjective pair. Two sets of four paragraphs each were then compiled
to make up two descriptive essays. The essays referred to two targets,
Donald and Ralph. The target essays are shown in the Appendix. The
bipolar trait adjectives corresponding to each paragraph are given in
parentheses at the end of each paragraph.
The two essays were presented to 100 Ohio State University undergraduates. Subjects were instructed to characterize the target portrayed
in each paragraph with a single word. Next, subjects made likability
ratings for each target and judged the extent to which each of the eight
bipolar adjective traits was referentially relevant to each target.
Two independent codersjudged the semantic similarity of each singleword characterization to either bipolar end of the corresponding trait
dimension. The coders were first provided with a list of synonyms for
each pair of the Donald and Ralph trait constructs. The coders were
told that the synonyms were not necessarily limiting and that in the case
that a subject's response was not included in the synonyms, they should
exercise their own judgment.
The Donald synonyms of the positive traits were the following: adventurous = daring, bold, courageous, brave; moral = ethical, principled,
conformist; independent = free, unconventional, individualistic; and
persistent = determined, persevering, steadfast. The Donald synonyms
of the negative traits were the following: reckless = careless, foolhardy,
rash; self-righteous = liar, superior, conceited; aloof = loner, distant,
unneighborly, unsociable; and stubborn = obstinate, unreasonable,
headstrong.
The Ralph synonyms of the positive traits were the following: idealistic = romanticist, optimist, dreamer; witty = amusing, humorous,
funny; cultivated = cultured, refined, educated; and thrifty = sparing,
provident. The Ralph synonyms of the negative traits were the following: unrealistic = silly, ignorant, disillusioned, utopian, stupid; sarcastic = critical, caustic, satiric, ironic; artificial = feigned, assumed,
pretentious, cunning; and stingy = miser, miserly, frugal. Synonyms
were selected from Allen's Synonyms and Antonyms (Vail Motter,
1972), Roget'S Thesaurus in Dictionary Form (Lewis, 1978), and Webster's New Collegiate Dictionary (1981). The coders agreed in 91% of
the cases and resolved disagreements through discussion.
The results (Table 1) highlight three points. First, the essays were indeed perceived as ambiguous: Subjects roughly were equally likely to
characterize the target using either bipolar adjective trait (or a synonym). Second, subjects regarded traits intended to describe Ralph as
relevant to Ralph and irrelevant to Donald, and regarded traits intended
to describe Donald as relevant to Donald and irrelevant to Ralph.
Third, subjects liked Donald better than Ralph--a finding that was replicated throughout this research.
Procedure
For half of the subjects, the priming task came first. Subjects were
greeted and escorted to the experimental room by Experimenter A (a
casually dressed man). Experimenter A introduced himself, called in
Experimenter B (a formally dressed man waiting outside the room), and
introduced him as a senior student who would use only a part of the
experimental hour to complete his honors thesis research. Experimenter A then stepped out of the room.
Experimenter B presented his research as a study concerned "with
people's ability to reorganize scrambled sentences and with their memory of those sentences." He informed subjects that they would have to
unscramble, memorize, and recall eight sentences.
There were four different sets of unscrambled sentences, corresponding to the four Valence X Applicability combinations. Each set included
eight sentences. Each sentence consisted of five words, only four of
which could make a sentence (e.g., adventurous is very door he; eye the
independent votes person; effort moral neck being requires). The order
of the sentences was randomized and kept constant across subjects.
In each set, each pair of sentences contained primes that were synonyms and that were relevant to the same paragraph. The positive prime
synonyms for the Donald paragraph were the following: adventurous/
daring, moral/ethical, independent/free, and persistent/determined.
The negative prime synonyms for the Donald paragraph were the following: reckless/careless, self-righteous/liar, aloof/loner, and stubborn/
obstinate. The positive prime synonyms for the Ralph paragraph were
the following: idealistic/romanticist, witty/amusing, cultivated/cultured, and thrifty/sparing. Finally, the negative prime synonyms for the
Ralph paragraph were the following: unrealistic/silly, sarcastic/critical,
artificial/feigned, and stingy/miser.
Subjects took approximately 3 min to complete the unscrambling
and memorization tasks. Next, subjects were given I rain to write down
on a sheet of paper as many states in the United States as they could
remember. Subsequently, subjects recalled the sentences for approximately 2 min. Subjects were then asked to return to the page containing
the scrambled sentences and to mark off the sentences they failed to
recall. At the conclusion of the session, Experimenter B thanked subjects, walked out, and thanked Experimenter A (who was waiting outside the room) out loud so that subjects could hear him.
The priming manipulation used in the present research was a combination of tasks introduced by Srull and Wyer (1979, 1980) and Higgins
et al. (1977). However, Srull and Wyer used a large number of unscrambled sentences and primed one trait construct only (e.g., either hostile
or kind), whereas the present research used a small number of unscrambled sentences and primed four trait constructs. Thus, in order to maximize the effectiveness of the present priming task, procedural characteristics of the Higgins et al. priming manipulation were borrowed. Higgins
et al. had subjects memorize the primes for a short time period. The
present research asked subjects to memorize the sentences after unscrambling them, to recall the sentences, and to note the sentences they
failed to recall. Moreover, the present research used traits as primes (as
did Higgins et al.) instead of using behaviors as primes (as did Srull &
Wyer). Finally, in contrast to both Higgins et al. and Srull and Wyer,
no filler sentences were used; that is, the eight unscrambled sentences
corresponded to the eight primes.
Care was taken so that subjects would not get suspicious of the relation between the priming task and the subsequent study. The dressing
style of the two experimenters was different, a rather elaborate procedural scenario was followed, and the experimental booklet in the priming task had a different color than did the experimental booklet in the
subsequent study (i.e., pink as opposed to white) and also had a different
typeface (i.e., printed with a dot matrix printer as opposed to a letterquality printer).
After the priming task, Experimenter A handed out a booklet. The
first page of the booklet informed subjects that they would read an essay
about an undergraduate student referred to as Ralph (Donald). Ralph
(Donald) was allegedly a member of a group of students who had allowed faculty and graduate students of the psychology department (Personality and Social Psychology Program) to study their personalities and
interpersonal relationships for the past year. Subjects were also told that
at a later point, they would have to describe Ralph (Donald) and communicate their impressions of him.
Subjects in the positive and negative recipient conditions were informed that their task was to describe and communicate their impressions of Donald (Ralph) to "another member of his group (Vincent P.),
so that this member will later be able to identify Donald (Ralph)" (patterned after Higgins & McCann, 1984). Subjects were told that their
descriptions and impressions of the target would be removed from the
booklet and be taken directly to the recipient for his use only. It was also
mentioned that the recipient happened to either like or dislike the target.
Subjects in the neutral recipient condition were instructed to describe
Ralph (Donald) and communicate their impressions of him. No infor-
PERSON IMPRESSIONS
401
Table 1
Results of Target Essay Construction
Percentage of subjects
Trait
Judging to
characterize
Ralph
Judging to
characterize
Donald
Judging as relevant
to Ralph
Judging as relevant
to Donald
91
34
81
36
79
36
86
25
82
16
80
53
75
46
91
30
Ralph paragraphs
Idealistic/unrealistic
Idealistic
Unrealistic
Witty/sarcastic
Witty
Sarcastic
Cultivated/artificial
Cultivated
Artificial
Thrifty/stingy
Thrifty
Stingy
60
3l
34
45
39
30
46
52
Donald paragraphs
Adventurous/reckless
Adventurous
Reckless
Moral/self-righteous
Moral
Self-righteous
Independent/aloof
Independent
Aloof
Persistent/stubborn
Persistent
Stubborn
46
4l
39
47
44
51
45
53
Note. Mean liking score for Donald was 4.27; mean liking score for Ralph was 3.90.
mation regarding the recipient's attitudes toward the target was provided. This dependent measure is referred to as communicated impres-
sions.
For the second half of the subjects, communication instructions preceded the priming task. Experimenter A informed subjects that a senior
student would step in anytime to do a 5-min experiment as a part of his
honors thesis. Next, Experimenter A had subjects go through the cover
page of the booklet, where the task of communicating the upcoming
target information to the recipient was described. This task was pretested to take about 2 min from the time subjects entered the room.
Instructions at the bottom of the cover page specified that subjects
should not turn to the next page until told to do so. After 2 min, Experimenter B knocked on the door and walked in. Experimenter A introduced Experimenter B and left the room. After the priming task was
over, Experimenter A came back and paced subjects through the target
information.
The rest of the procedure was common to all subjects. The target
essay appeared on the second page of the booklet, and subjects communicated their impressions of the target on the third page of the booklet.
Subjects were subsequently provided with a list of the states in the
United States and were asked to write down the capitals. After 2 min,
subjects were told that although the task of describing the target and
communicating the impressions of him to Vincent P. was over, the experimenter would still like them to provide some additional information. This would be for the use of the experimenter only, so that he
could examine "other aspects of the judgmental task." Subjects were
requested to write down as fully as they could what sort of person they
privately thought the target was. This is referred to as theprivate impressions task.
Next, subjects were presented with the target essay, paragraph by
paragraph, and were asked to characterize with a single word the kind
of person depicted by each paragraph description. This is the wordgeneration task. The priming procedure was hypothesized to activate trait
categories, not just trait words. If subjects repeatedly used the exact
primes to characterize target behaviors, construct accessibility effects
would have been shown to simply be a lexical phenomenon. However,
if subjects used prime synonyms to label target behaviors, it would be
conclusively demonstrated that the priming technique rendered an entire trait category accessible, not just a word (cf. Higgins et al., 1977).
Subjects subsequently went through three manipulation checks,
probing into the purpose of the experiment, suspicions of the relatedness between the priming task and the communication tasks, and memory of recipient attitudes toward the target. Finally, subjects were asked
to recall the unscrambled sentences for a second time and then were
debriefed, thanked for their participation, and excused.
Coding
Subjects' communicated and private impressions were coded by two
independent coders. Protocols were presented to the coders in random
order, each coder receiving a different random order. (This was the case
throughout the reported research.)
The coders were first provided with a list of synonyms for each pair
of the Donald and Ralph primes. The list of synonyms was the same as
402
CONSTANTINE SEDIKIDES
the ones used by the coders involved in the construction of the target
essays. The coders were told that they should exercise their own judgment in case subjects' responses could not be coded in line with the
synonyms. The coders were instructed to evaluate subjects' impressions
of Donald on the basis of the Donald-applicable primes and to evaluate
subjects' impressions of Ralph on the basis of the Ralph-applicable
primes. The coders were unaware of the valenceof the primes.
The coders broke down subjects' impressions into four paragraphs
and then rated each paragraph on a scale ranging from 1 (distorted to
be the same as applicable negative primes or synonyms) through 4 (ambiguous, undistorted, or same as the original paragraph) to 7 (distorted
to he the same as applicable positive primes or synonyms). The mean of
the four paragraphs was calculated for each coder; the mean of the two
coder means was entered into the analyses. Whenevera paragraph was
missing(in approximately 11% of the cases), the mean was based on the
remaining three scores. Coder intercorrelations were high throughout
the reported research (rs > .81, ps < .001).
The word-generationtask was next. The coders noted the percentage
of the trait terms generated by subjects to characterize the target that
were the same as the primes. Finally, the coders coded subjects' responses on the manipulation checks.
Results and Discussion
For an answer to the question about the purpose of the study
to qualify as correct, the answer needed to include a mention of
the relation between the sentence-unscrambling task and the
subsequent processing of the target essay (e.g., whether the sentence-unscramblingtask influenced the interpretation of the essay). A gist criterion was used. None of the subjects correctly
guessed the purpose of the study. Furthermore, 98% of the subjects correctly remembered the recipient's attitude toward the
target.
Subjects' memories of the primes was examined, using the
second prime recall only. Subjects recalled an average of 6.3 out
of 8 primes, indicating that they had committed the primes to
memory. No prime recall differences among experimental conditions were observed.
Communicated Impressions
The targets were described most favorably to the positive recipient (M = 5.41 ), followed by the neutral recipient (M = 4.10)
and the negative recipient (M = 2.61), F(2, 144) = 142.76, p <
.000001. (The three means significantly differed from one another, as indicated by the analyses that follow.) Recipient characteristics affected subjects' impressions, a finding that replicates past research (e.g., Higgins & Rholes, 1978; Higgins &
McCann, 1984).
Of crucial importance was whether communication goals
moderated construct accessibility effects. The Valence X Applicability X Recipient interaction was significant, F(2, 144) =
9.11, p < .0002 (Table 2). To pinpoint the locus of the effect, I
broke down the interaction into a number of lower order interactions that are reported later in this article. It should be mentioned at this point that the Valence X Applicability X Recipient X Order interaction was not significant, F(2, 144) = 0.66,
p < .5 l, indicating that the pattern of results was not affected
by the temporal order of the priming and communication tasks.
Neutral recipient. An analysis of variance (ANOVA) was carried out on the neutral recipient condition only. The Valence x
Applicability interaction was significant, F(l, 48) = 35.72, p <
Table 2
Means on Communicated Impressions as a Function oJ"
Valence, Applicability, and Recipient in Study I
Prime
Positive
recipient
Neutral
recipient
Negative
recipient
Applicable
Positive
Negative
5.31
5.29
5.84
2.74
2.71
2.44
Inapplicable
Positive
Negative
5.67
5.37
4.20
3.62
2.43
2.88
.000001. Positive applicable primes caused positive distortion
of subjects' communicated impressions, and negative applicable primes caused negative distortion of communicated impressions. In contrast, valence of inapplicable primes did not cause
differential distortion of communicated impressions. These results extend past construct accessibility effects (e.g., Higgins et
al., 1977; Sinclair, Mark, & Shotland, 1987) by showing that
semantic priming can affect not only the words subjects choose
to characterize the target but also the content of person impressions that are communicated in natural language.
Positive versus neutral recipient. An ANOVA on the positive
and neutral recipient conditions revealed a reliable Valence x
Applicability X Recipient interaction, F(l, 96) = 22.28, p <
.000001. It was only in the case of the neutral recipient that
applicable positive and negative primes distorted subjects' communicated impressions. Primes were ineffective in the case of
the positive recipient.
Negative versus neutral recipient. An ANOVAon the negative
and neutral recipient conditions disclosed a reliable Valence x
Applicability x Recipient interaction, F(l, 96) = 6.60, p < .01.
Only in the case of the neutral recipient did applicable positive
and negative primes distort subjects' communicated impressions. Primes were ineffective in the case of the negative recipient.
Positive versus negative recipient. An ANOVAon the positive
and negative recipient conditions yielded a nonsignificant Valence x Applicability X Recipient interaction, F(1, 96) = 2.08,
p < . 15, and a nonsignificant Valence X Applicability interaction, F(1, 96) = 0.41, p < .52. Construct accessibility effects
were eliminated when the recipient's attitude toward the target
became known to the subjects. Communication goal effects
overrode construct accessibility effects.
Private I m p r e s s i o n s
Subjects formed a more positive impression of the targets after having described them to the positive recipient (M = 5.28),
followed by the neutral recipient (M = 4.00) and the negative
recipient (M = 2.67), F(2, 144) = 154.62, p < .000001. (The
three means differed significantly from each other, as revealed
by the analyses reported later.) Subjects' private impressions of
the targets were shaped by their communicated impressions.
The Valence x Applicability X Recipient interaction was significant, F(2, 144) = 6.62, p < .001 (Table 3). Separate analyses
that evaluate the interaction are reported in the next four para-
403
PERSON IMPRESSIONS
Table 3
Means on Private Impressions as a Function of Valence,
Applicability, and Recipient in Study 1
Positive
recipient
Neutral
recipient
Negative
recipient
Applicable
Positive
Negative
5.40
5.08
5.55
2.80
2.78
3.53
Inapplicable
Positive
Negative
5.37
5.26
4.05
3.58
2.28
2.91
Prime
graphs. The Valence X Applicability X Recipient × Order interaction was again unreliable, F(2, 144) = 0.07, p < .93, indicating that the order of the priming and communication tasks did
not interfere with subjects' private impressions.
Neutral recipient. The Valence X Applicability interaction,
F( I, 48) = 30.43, p < .000001, revealed that subjects' own impressions of the targets were positively distorted in the condition
of positive applicable primes and were negatively distorted in
the condition of negative applicable primes. Valence of inapplicable primes did not account for differences in subjects' distortions of their impressions.
Positive versus neutral recipient. The Valence X Applicability x Recipient interaction was significant, F ( l , 96) = 12.74,
p < .006. Construct accessibility effects were operative only in
the case of the neutral recipient.
Negative versus neutral recipient. The Valence × Applicability X Recipient interaction was significant, F ( l , 96) = 6.82, p <
.01. Construct accessibility effects were present only in the case
of the neutral recipient.
Positive versus negative recipient. The Valence x Applicability x Recipient interaction was not significant, F(1, 96) = 0.67,
p < .41. The Valence X Applicability interaction was not significant either, F ( I , 96) = 2.36, p < . 12. Construct accessibility
effects were not detectable. Communication goal effects predominated over construct accessibility effects in subjects' private impressions of the targets.
Word Generation
Only 16.7% of the words that subjects used to characterize
the target were the same as the primes. The obtained construct
accessibility effects were not simply a lexical phenomenon.
Study 2
Study 1 established that when no information about recipient
characteristics is given to subjects, construct accessibility effects
take precedence over communication goal effects; when information about recipient characteristics becomes available to
subjects, communication goal effects take precedence over construct accessibility effects. Communication goals have a strong
distortive impact on person impressions. This impact, in turn,
colors subjects' own characterizations of the target.
The impact, however, of communication goals on person impressions needs to be qualified. Communication goals exerted
their influence before the encoding of target information. It is
likely that when communication goals are introduced after the
encoding of target information, their effects are overpowered by
construct accessibility effects.
In Study 2, the recipient's attitudes became known (to half of
the subjects) after encoding. Thus, Study 2 addressed the issue
of whether communication goals can trigger recoding of target
information.
Method
Subjects were 288 Ohio State University undergraduates. The procedure and design of the study was identical to Study 1 with one exception:
the inclusion of the location of communication instructions variable.
For half of the subjects, communication instructions were administered
before the encoding of target information, a conceptual replication of
Study I. For the remaining half of the subjects, communication instructions were delivered after the encoding of target information.
Results
No subject guessed the purpose of the study or suspected the
relation between the two parts of the study. Subjects recalled an
average of 6.2 sentences. Moreover, 98% of the subjects correctly recalled the recipient's attitudes.
C o m m u n i c a t e d Impressions
The targets were described most favorably to the positive recipient (M = 4.25), followed by the neutral recipient (M--- 3.76)
and the negative recipient (M = 3.11), F(2, 240) = 10.76, p <
.000001. (The negative recipient mean significantly differed
from both the positive and the neutral recipient means, as the
analyses that follow reveal.) Subjects' communicated impressions were in line with their recipient's attitudes, a replication
of the relevant finding of Study 1.
The critical Valence x Applicability x Recipient interaction
was not significant, F(2, 240) = 1.77, p < .17, although the
means were in the expected direction (see Table 4). Most important, the Valence X Applicability × Recipient X Location interaction was unreliable, F(2, 240) = 0.27, p < .76. The moderation of communication goal effects on construct accessibility
effects was not a function of whether the communication instructions were delivered before or after encoding.
Neutral recipient. An ANOVA on the neutral recipient condition revealed a significant Valence x Applicability interaction,
F( 1, 80) = 5.81, p < .01. Positive applicable primes caused positive distortion of communicated impressions, whereas negative
Table 4
Means on Communicated Impressions as a Function of
Valence, Applicability, and Recipient in Study 2
Positive
recipient
Neutral
recipient
Negative
recipient
Applicable
Positive
Negative
4.63
4.76
4.51
2.96
3.14
3.37
Inapplicable
Positive
Negative
3.22
4.42
3.72
3.88
2.92
3.02
Prime
404
CONSTANTINE SEDIKIDES
Table 5
Means on Private Impressions as a Function of Valence,
Applicability, and Recipient in Study 2
Positive
recipient
Neutral
recipient
Negative
recipient
Applicable
Positive
Negative
3.71
4.02
4.19
2.42
2.80
3.21
Inapplicable
Positive
Negative
2.90
3.28
3.59
3.86
2.67
2.94
Prime
applicable primes caused negative distortion of communicated
impressions. In contrast, valence of inapplicable primes did not
cause differential distortion of communicated impressions.
Construct accessibility effects were operative when no knowledge about recipient attitudes was provided.
The Valence X Applicability X Location ANOVA was computed to answer the question of whether the introduction of
communication goals after encoding has a distortive impact on
social information. A nonsignificant interaction, F(1, 80) =
0.97, p < .32, showed that construct accessibility effects overpowered communication goal effects regardless of whether
communication instructions were delivered before or after encoding.
Positive versus neutral recipient. An ANOVA on the positive
and neutral recipient conditions produced a nonsignificant Valence x Applicability X Recipient interaction, F(1, 160) = 0.45,
p < .50. The relevant finding of Study 1--that construct accessibility effects were evident in the neutral but not in the positive
recipient case--was not replicated. The effect did not interact
with location, F(I, 160) = 0.01, p < .90.
Negative versus neutral recipient. An ANOVA on the negative
and neutral recipient conditions revealed a marginally significant Valence x Applicability X Recipient interaction, F(1,
160) = 3.43, p < .06, suggesting that only in the case of the
neutral recipient did the applicable positive and negative
primes distort subjects' communicated impressions. The
primes were ineffective in the case of the negative recipient.
This finding held, regardless of temporal order of the priming
task and communication instructions, F(1, 160) = 0.47,
p < .49.
Positive versus negative recipient. An ANOVAon the positive
and negative recipient conditions yielded a nonsignificant Valence X Applicability x Recipient interaction, F(1,160) = 1.42,
p < .23, and a nonsignificant Valence x Applicability interaction, F(1, 160) = 0.89, p < .34. Construct accessibility effects
were not detectable. Most important, the Valence x Applicability X Recipient × Location interaction was unreliable, F(1,
160) = 0.33, p < .56. Communication goal effects overrode construct accessibility effects even when communication instructions were delivered after the priming task. Communication
goals led to reinterpretation of target information.
Private Impressions
Two subjects were excluded from the analyses because of their
failure to provide usable data. The targets made a more favor-
able impression after being described to either the positive recipient (M = 3.47) or the neutral recipient (M = 3.50), as compared with the negative recipient (M = 2.89), F(1,238) = 3.14,
p < .04. The negative recipient mean differed significantly from
either the positive or the neutral recipient mean, as the analyses
that are reported later in this article revealed.
The Valence X Applicability X Recipient interaction was
marginally significant, F(2,238) = 2.44, p < .08 (Table 5). Separate analyses are reported later to evaluate this interaction.
The Valence X Applicability X Recipient X Location interaction was not significant, F(2, 238) = 0.44, p < .64. Order of the
priming task and communication instructions did not differentially affect subjects' private impressions.
Neutral recipient. Subjects' private impressions of the targets
were positively distorted under positive applicable primes and
were negatively distorted under negative applicable primes. Valence of inapplicable primes did not account for differences in
subjects' distortions of their own impressions, F(1, 79) = 7.54,
p < .007.
Temporal location of communication goals did not influence
subjects' evaluations of the target, as reflected in the nonsignificant Valence x Applicability X Location interaction, F(I,
79) = 0.94, p < .33. Retrospective encoding did not affect subjects' own impressions.
Positive versus neutral recipient. The Valence X Applicability x Recipient interaction was marginally significant, F(1,
159) = 3.06, p < .08. Construct accessibility effects tended to
be evident in the neutral but not in the positive recipient
case. The effect did not interact with location, F( 1, 159) = 0.00,
p < .95.
Negative versus neutral recipient. The Valence x Applicability x Recipient interaction was significant, F(1, 158) = 4.52,
p < .03. Construct accessibility effects were present in the neutral but not in the negative recipient case. Again, the effect did
not interact with location, F(I, 158) = 0.69, p < .40.
Positive versus negative recipient. The Valence X Applicability X Recipient interaction was not significant, F(1, 159) =
0.03, p < .85. The Valence x Applicability interaction was not
significant either, F(1, 159) = 0.00, p < .94. Communication
goal effects overpowered construct accessibility effects. Furthermore, temporal order of the priming task did not have any detectable effect on subjects' private impressions: The interaction
involving location was unreliable, F(1, 159) = 0.68, p < .41.
Only 15.8% of the words subjects used to characterize the
target were the same as the primes. Construct accessibility
effects, when obtained, were not simply a lexical phenomenon.
General Discussion
Models intended to explain construct accessibility effects
(Higgins & King, 1981; Wyer & Srull, 1981) postulate that new
social information is encoded or interpreted in line with the
most accessible and applicable stored constructs.
One line of communication research (e.g., Clark, 1985;
Grice, 1975; Higgins, 1981) highlights the functional importance of interpersonal communication in everyday life. People
communicate with one another in order to achieve various interpersonal goals (McCann & Higgins, 1988). A precondition
for the attainment of interpersonal goals is the establishment of
rapport, which in turn presupposes taking into account recipi-
PERSON IMPRESSIONS
ents' characteristics. The line of research holds that information about recipients' idiosyncrasies plays a crucial role in the
interpretation and communication of social information. Another line of communication research (Zajonc, 1960) posits
that communicators resist closing in on target impressions before acquiring complete knowledge of the communicative situation. This line of research also stresses the role of communication goals in determining the formation of person impressions.
The present results were partially consistent with models of
construct accessibility effects. People used the most accessible
and applicable constructs to interpret and communicate social
information when knowledge about recipients' characteristics
was unavailable. Furthermore, construct accessibility effects
held, regardless of preencoding order of the priming and communication tasks and regardless of whether the communication
task was initiated before or after encoding of target information.
In contrast, the remainder of the findings were consistent
with research on communication goal effects. Subjects' interpretation and communication of social information were influenced by communication goals governing social interaction
(i.e., being in coordination with recipients). It did not matter
whether communication instructions were introduced before or
after the priming task or before or after the encoding of target
information.
Construct accessibility researchers have called for research
on the conditions under which social information will be recoded (Srull & Wyer, 1980). Study 2 delineated one such condition: Social information will be recoded when subjects become
aware of recipient characteristics prior to communication.
405
1982; Lingle & Ostrom, 1979). Furthermore, people's impressions of a target become more polarized over time (Higgins et
al., 1977), especially when they are communicated (Higgins &
McCann, 1984; Higgins & Rholes, 1978). Delayed target impressions would, if anything, be more consistent with previously reported impressions.
In the private impressions task, subjects were given an opportunity to express their honest evaluations of the target. If subjects' private impressions of the target were different from their
communicated impressions, subjects should have indicated so.
The present results complement the emerging view that people
can recode stored information (e.g., Hirt & Castellan, 1988).
Consequencesfor Models of Construct
Accessibility Effects
To account for the present findings, both mechanistic and excitation transmission models would need to take into consideration communication goal effects. Mechanistic models would
need to postulate that constructs activated through communication goals (task-relevant constructs) can top fortuitously activated constructs (task-irrelevant constructs). Excitation transmission models would need to posit that communication-relevant constructs can be excited to a threshold that surpasses the
threshold of constructs activated through fortuitous experiences.
Recoding, Impression Management, or Response Bias?
Memory for Primes and Assimilation
Versus Contrast Effects
A recoding explanation of the results of Study 2 has been
offered. Communication goals became salient to subjects after
encoding. At that moment, subjects selectively reviewed target
information, focusing on (and subsequently transmitting) only
those aspects of the information that would facilitate communication with the recipient. Subjects subsequently believed their
own message ("saying is believing"; Higgins & Rholes, 1978).
There are at least two alternative interpretations of the results. An impression management interpretation maintains
that subjects were concerned only with creating a favorable impression both on the recipient (hence, subjects' efforts to coordinate their evaluations of the target with the recipient's evaluation) and on the experimenter (hence, the consistency over time
between communicated and private impressions). The plausibility of an impression management explanation, though, is
challenged by two design features. First, subjects' responses
were anonymous. Second, subjects were told that their communicated impressions would be taken directly to the recipient,
whereas their private impressions would be separately examined by the experimenter.
A response bias interpretation of the obtained results posits
that subjects simply found it easy to repeat their communicated
impressions of the target in the private impressions task. Had
the delay between the communicated impressions and the private impressions tasks been longer, the results might have been
different. However, people tend to rely on their past judgments
or impressions, rather than on their stored representations, for
subsequent judgments of a target (Carlston, 1980; Higgins et al.,
The present findings have implications for the debate regarding the circumstances under which assimilation versus contrast
effects are obtained. Lombardi, Higgins, and Bargh (1987; see
also Dark, 1988; Martin, 1986) had subjects recall the primes
after the completion of the priming task. Good memory for
primes (that is, conscious awareness of the primes) was associated with contrast effects, whereas poor memory for primes was
associated with assimilation effects.
The results of the present research are incompatible with
Lombardi et al's (1987) results. Subjects in the present research
manifested good recall for the primes and concurrent assimilation effects. Assimilation effects are surprising, because semantic processing of the primes was involved. Subjects unscrambled the priming sentences, memorized them for meaning, and
recalled them. Clearly, subjects became involved in the priming
task and processed the priming sentences deeply (cf. Craik &
Lockhart, 1972).
It is not clear how to interpret the discrepancy between the
present findings and Lombardi et al.'s (1987) findings. Lombardi et al. maintained that consciousness of primes may lead
to more flexible and differentiated processing of target information than may absence of consciousness of primes. That is, consciousness of primes may lead to both assimilation and contrast
effects, with the determining factor being aspects of the experimental procedure or task. Given the differences between the
present studies and Lombardi et al.'s studies in terms of the
priming task, the experimental instructions, and the response
output format, additional research is needed to clarify the issue.
406
CONSTANTINE SEDIKIDES
Relative Influence o f Construct Accessibility and
C o m m u n i c a t i o n Goal Effects
The present research showed that communication goal
effects can override construct accessibility effects. However, the
conclusion that construct accessibility effects will invariably
vanish or get attenuated in the presence of knowledge about
recipient characteristics is unwarranted. There may be cases in
which construct accessibility effects are present even when recipient attitudes become known to subjects. The reported research was confined to a single research paradigm. Different
research paradigms may produce different results. For example, repeated priming, as opposed to single-time priming, may
lead to an attenuation of communication goal effects.
Alternatively, there may exist other communication-related
factors, besides recipient characteristics, that block out construct accessibility effects. Communicators engage in mental activity in order to understand the recipients' cognitive content.
For example, communicators will routinely modify their message as a function of whether the information they supply to
their recipients is new or old (Clark & Haviland, 1977). When
providing new information, communicators may overly engage
in detail in their effort to best explain informational intricacies
to recipients. As a result, the impact of accessible constructs is
likely to be attenuated (see Strack et al., 1988). Furthermore,
communicators may be involved in role taking as a means of
facilitating communication. Assuming a new role often means
activating an entirely different set of stored knowledge. The impact of accessible constructs is again likely to be attenuated.
Finally, the influence of accessible constructs may be d i m i n ished in cases in which face-to-face communication with a recipient takes place.
Conclusion
Communication can best be understood in the context of social cognition, because social cognition is the basis for communication. Social cognition--which involves knowledge about
the personal characteristics, needs, and expectations of cointeractants--is essential in formulating the appropriate message.
Furthermore, communication relies on social knowledge, because it frequently involves the exchange of information about
others (Clark, 1985; Higgins, 1981 ).
Social cognition also can best be understood in the context
of communication. The communicative message is the product
of intense cognitive activity taking place in the communicator's
mind, with cognitive activity being equivalent to transformations of cognitive structures. Moreover, emitting a message can
modify the communicator's own knowledge about the recipient. By studying the communication processes, one can advance understanding of the nature and workings of cognitive
structures (Kraut & Higgins, 1984).
Forgas (1981) considered the communicative situation to be
the very essence of social cognition. Markus and Zajonc (1985)
maintained that one-way information-processing flowcharts do
not give justice to the communicative situation. As these authors put it,
it is likely that in the near future the major new method of studying
social cognition and of cognition in general will be the dialogue,
supplementing the paradigm of recognition memory and reaction
time. Individual subjects in interaction, each asking the other questions and responding, may disclose a great deal of content and
structure of their own cognitions and help reveal the cognitions of
the other. (Markus & Zajonc, 1985, p. 212)
References
Anderson, N. H. (1974). Cognitive algebra: Integration theory applied
to social attribution. In L. Berkowitz (Ed.), Advances in experimental
socialpsychology (Vol. 7, pp. 1-101 ). Orlando, FL: Academic Press.
Bargh, J. A., & Pietromonaco, P. (1982). Automatic information processing and social perception: The influence of trait information presented outside of conscious awareness on impression formation.
Journal of Personality and Social Psychology, 43, 437-449.
Brewer, M. B. (1988). A dual process model of impression formation.
In T. K. Srull & R. S. Wyer, Jr. (Eds.), Advances in social cognition
(Vol. 1, pp. 1-36). Hillsdale, N J: Erlbaum.
Bruner, J. S. (1957). On perceptual readiness. PsychologicalReview, 64,
123-152.
Carlston, D. E. (1980). Events, inferences, and impression formation.
In R. Hastie, T. M. Ostrom, E. B. Ebbesen, R. S. Wyer, Jr., D. L.
Hamilton, & D. E. Carlston (Eds.), Person memory: The cognitive
basis ofsocialperception (pp. 89-119). Hillsdale, N J: Erlbaum.
Chafe, W. L. (1974). Language and consciousness. Language. 50, 111133.
Clark, H. H. (1985). Language use and language users. In G. Lindzey
& E. Aronson (Eds.), The handbook of social psychology (Vol. 2, pp.
179-231 ). New York: Random House.
Clark, H. H., & Haviland, S. (1977). Comprehension and the given-new
contract. In R. Freedle (Ed.), Discourse production and comprehension (pp. 1-40). Norwood, N J: Ablex.
Craik, E I., & Lockhart, R. S. (1972). Levels of processing: A framework
for memory research. Journal of Verbal Learning and VerbalBehavior, 11, 671-684.
Dark, V. J. (1988). Semantic priming, prime reportability, and retroactive priming are interdependent. Memory & Cognition, 16, 299-308.
Donohew, L., Sypher, H. E., & Higgins, E. T. (Eds.). (1988). Communication, social cognition, and affect. Hillsdale, N J: Erlbaum.
Forbach, G. B., Stanners, R. E, & Hochhaus, L. (1974). Repetition and
practice effects in a lexical decision task. Memory & Cognition, 2,
337-339.
Forgas, J. P. (1981). Social cognition: Perceptives on everyday understanding. Orlando, FL: Academic Press.
Grice, H. ( 1975). Logic and conversation. In P. Cole & J. Morgan (Eds.),
Syntax and semantics: Speech acts (Vol. 3, pp. 41-58). Orlando, FL:
Academic Press.
Herr, P. M. (1986). Consequences of priming: Judgment and behavior.
Journal of Personality and Social Psychology, 51, 1106-1115.
Higgins, E. T. (1981). The "communication game": Implications for
social cognition and persuasion. In E. T. Higgins, C. P. Herman, &
M. P. Zanna (Eds.), Social Cognition: The Ontario Symposium (Vol.
1, pp. 343-392). Hillsdale, NJ: Erlbaum.
Higgins, E. T., & King, G. A. (1981). Accessibility of social constructs:
Information processing consequences of individual and contextual
variability. In N. Cantor & J. E Kihlstrom (Eds.), Personality, cognition, and social interaction (pp. 69-121 ). HiUsdale, N J: Erlbaum.
Higgins, E. T., & Lurie, L. (1983). Context, categorization, and recall:
The "change-of-standard" effect. Cognitive Psychology, 15. 525-547.
Higgins, E. T., & McCann, C. D. (1984). Social encoding and subsequent attitudes, impressions, and memory: "Context-driven" and
motivational aspects of processing. Journal of Personality and Social
Psychology, 47, 26-39.
Higgins, E. T., McCann, C. D., & Fondacaro, R. (1982). The "communication game": Goal-directed encoding and cognitive consequences.
Social Cognition, 1, 21-37.
Higgins, E. T., & Rholes, W. (1978). "Saying is believing": Effects of
PERSON IMPRESSIONS
message modification of memory and liking for the person described.
Journal of Experimental Social Psychology, 14, 363-378.
Higgins, E. T., Rholes, W. S., & Jones, C. R. (1977). Category accessibility and impression formation. Journal of Experimental Social Psychology 13, 141-154.
Higgins, E. T., & Stangor, C. (1988). Context-driven social judgment
and memory: When "behavior engulfs the field" in reconstructive
memory. In D. Bar-Tal & A. W. Kruglanski (Eds.), The socialpsychology of knowledge (pp. 262-298). New York: Cambridge University
Press.
Hilton, D. J., & Slugoski, B. R. (1986). Knowledge-based causal attributions: The abnormal conditions focus model. Psychological Review,
93, 75-88.
Hirt, E., & Castellan, N. J. (1988). Probability and category redefinition
in the fault tree paradigm. Journal of Experimental Psychology:Human Perceptionand Performance, 14, 122-131.
Holtgraves, T., Srull, T. K., & Socall, D. (1989). Conversation memory:
The effects of speaker status on memory for the assertiveness of conversation remarks. Journal of Personality and Social Psychology, 56,
149-160.
Jones, E. E., & Davis, K. E. (1965). From acts to dispositions: The attribution process in person perception. In L. Berkowitz (Ed.), Advances
in experimental socialpsychology (Vol. 2, pp. 219-266). Orlando, FL:
Academic Press.
Kraut, R. E., & Higgins, E. T. (1984). Communication and social cognition. In R. S. Wyer, Jr., & T. K. Srull (Eds.), Handbook of social cognition (Vol. 3, pp. 87-127). Hillsdale, N J: Erlbaum.
Lewis, N. (Ed.). (1978). Roget's thesaurus in dictionaryform. New York:
Berkeley.
Lingle, J. H., & Ostrom, T. M. (1979). Retrieval selectivity in memorybased judgments. Journal of Personality and Social Psychology, 37,
180-194.
Lombardi, W. L., Higgins, E. T., & Bargh, J. A. (1987). The role of
consciousness in priming effects on categorization: Assimilation versus contrast as a function of awareness of the priming task. Personality and Social Psychology Bulletin, 13, 411-429.
Manis, M., Cornell, S., & Moore, J. (1974). Transmission of attituderelevant information through a communication chain. Journal of
Personality and Social Psychology, 30, 81-94.
Markus, H., & Zajonc, R. B. (1985). The cognitive perspective in social
psychology. In G. Lindzey & E. Aronson (Eds.), The handbook of
socialpsychology (Vol. l, pp. 137-230). New York: Random House.
Martin, L. L. (1986). Set/reset: Use and disuse of concepts in impression formation. Journal of Personality and Social Psychology, 51,
493-504.
McCann, C. D., & Hancock, R. D. (1983). Self-monitoring in communicative interactions: Social cognitive consequences of goal-directed
message modification. Journal of Experimental Social Psychology,
19, 109-121.
McCann, C. D., & Higgins, E. T. (1988). Motivation and affect in inter-
407
personal relations: The role of personal orientations and discrepancies. In L. Donohew, H. E. Sypher, & E. T. Higgins (Eds.), Communication, social cognition, and affect (pp. 53-79). Hillsdale, NJ: Erlbaum.
Newtson, D., & Czerlinsky, T. (1974). Adjustment of attitude communications for contrasts by extreme audiences. JournalofPersonality and
Social Psychology, 30, 829-837.
Peabody, D. (1967). Trait inferences: Evaluative and descriptive aspects
[Monograph]. Journal of Personality and Social Psychology, 7(4,
Whole No. 644).
Reder, L. M. (1983). What kind of pitcher can a catcher fill? Effects of
priming in sentence comprehension. Journal of VerbalLearning and
VerbalBehavior, 22, 189-202.
Sinclair, R. C., Mark, M. M., & Shotland, R. L. (1987). Construct accessibility and generalizability across response categories. Personality
and Social Psychology Bulletin, 13, 239-252.
Smith, E. R. (1984). Model of social inference processes. Psychological
Review, 91, 392-413.
Smith, E. R., & Branscombe, N. R. (1987). Procedurally mediated social inferences: The case of category accessibility effects. Journal of
Experimental Social Psychology, 23, 361-382.
Smith, E. R., & Branscombe, N. R. (1988). Category accessibility as
implicit memory. Journal of Experimental Social Psychology,, 24,
490-504.
Srull, T. K., & Wyer, R. S., Jr. (1979). The role of category accessibility
in the interpretation of information about persons: Some determinants and implications. Journal of Personality and Social Psychology,
37, 1660-1672.
Srull, T. K., & Wyer, R. S., Jr. (1980). Category accessibility and social
perception: Some implications for the study of person memory and
interpersonal judgments. Journal of Personality and Social Psychology, 38, 841-856.
Strack, E, Martin, L. L., & Schwarz, N. (1988). Priming and communication: Social determinants of information use in judgments of life
satisfaction. European Journal of Social Psychology, 18, 429--442.
Vail Motter, T. H. (Ed.). (1972). Allen "ssynonyms and antonyms (rev. &
enlarged ed.). New York: Barnes & Noble.
Webster's new collegiate dictionary (8th ed.). ( 1981). Springfield, MA:
Merriam-Webster.
Wyer, R. S., Jr., & Srull, T. K. ( 1981 ). Category accessibility: Some theoretical and empirical issues concerning the processing of social stimulus information. In E. T. Higgins, C. P. Herman, & M. P. Zanna
(Eds.), Social Cognition: The Ontario Symposium (Vol. 1, pp. 161197). Hillsdale, N J: Erlbaum.
Wyer, R. S., Jr., & Srull, T. K. (1986). Human cognition in its social
context. PsychologicalReview, 93, 322-359.
Zajonc, R. B. (1960). The process of cognitive tuning of communication. Journal of Abnormal and Social Psychology, 61, 159-167.
Zajonc, R. B., & Adelmann, P. K. (1987). Cognition and communication: A story of missed opportunities. SocialSciencelnformation, 26,
3-30.
(Appendix follows on next page )
408
CONSTANTINE SEDIKIDES
Appendix
Target Essays Used in Studies 1 and 2
Ralph
DonaM
Ralph has his own ideas about bow people should live. He feels that the
world can be a much better place to live if everyone practices transcendental meditation. He envisions a world without hunger, poverty, or
crime. He believes that once such a world is established, all political
leaders can then step down from their posts, and all people, regardless
of age, race, or nationality, will live in harmony together. (Unrealistic,
Donald spends a great amount of his time in search of what he likes
to call excitement. He has already climbed Mt. McKinley, done some
skydiving, shot the Colorado rapids in a Kayak, driven in a demolition
derby, and piloted a jet-powered boat--without knowing much about
boats. He has been injured, and even risked death, a number of times.
Ideal&tic)
Donald has his own standards of behaving. As a student he would tell
on fellow classmates whom he saw break school rules, like cheating on
tests. In fact, he claimed to his friends that never once in his life he had
thought about cheating. (Moral Self-righteous)
Other than business engagements, Donald's contacts with people are
surprisingly limited. He feels he doesn't really need to rely on anyone.
A lot of people enjoy Ralph's humor. His is in the habit of making jokes
out of the blue. Often times in parties his humor is quick to address
the faults that people have or the mistakes that they have made. ( Witty,
Sarcastic)
Ralph recently started making attempts to keep up to date with cultural
knowledge. He read a book about Europe, sat in a music appreciation
workshop, and eats at fashionable ethnic restaurants. When being with
friends, he often talks at length about foreign culture and art. (Culti-
vated, Artificial)
In order to improve his life Ralph tries to save money. He uses coupons,
buys things on sale, and avoids donating money to charity or lending
money to friends. (Thrifty, Stingy)
(Adventurous, Reckless)
(Aloof Independent)
Once Donald makes up his mind to do something it is as good as done
no matter how long it might take or how difficult the going might be.
Only rarely does he change his mind even when it might be better if he
had. (Persistent, Stubborn)
Received M a r c h 17, 1989
Revision received October 25, 1989
Accepted O c t o b e r 26, 1989 •